Assessment technique for acnetreatments based on statisticalparameters of skin thermal images

J. Alfredo Padilla-MedinaFrancisco León-OrdoñezJuan Prado-OlivarezNoe Vela-AguirreAgustin Ramírez-AgundisJavier Díaz-Carmona

Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Biomedical-Optics on 21 Jul 2020Terms of Use: https://www.spiedigitallibrary.org/terms-of-use

Assessment technique for acne treatments basedon statistical parameters of skin thermal images

J. Alfredo Padilla-Medina, Francisco León-Ordoñez, Juan Prado-Olivarez, Noe Vela-Aguirre,Agustin Ramírez-Agundis, and Javier Díaz-Carmona*Technological Institute of Celaya, Electronics Engineering Department, Av Tecnologico y G. Cubas s/n, Celaya Gto 38010, Mexico

Abstract. Acne vulgaris as an inflammatory disease, with an excessive production of subdermal fat, modifiesthe dynamics of the bloodstream, and consequently temperature, on the affected skin zone. A high percentageof this heat interchange is manifested as electromagnetic radiation with far-infrared wavelengths, which can becaptured through a thermal imaging camera. A technique based on thermal image analysis for efficiencyassessment in acne vulgaris is described. The procedure is based on computing statistical parameters ofthermal images captured from the affected skin zone being attended by an acne treatment. The proposedtechnique was used to determine the skin thermal behavior according to acne severity levels in differentacne treatment stages. Infrared images of acne skin zones on eight patients, diagnosed with acne vulgarisand attended by one specific acne treatment, were weekly registered during 11 weeks. The infrared imageswere captured until no more improvement in affected zones was detected. The obtained results suggest adirect relationship between the used statistical parameters, particularly first- and second-order statistics,and the acne vulgaris severity level on the affected zones. © The Authors. Published by SPIE under a Creative Commons

Attribution 3.0 Unported License. Distribution or reproduction of this work in whole or in part requires full attribution of the original publication, including

its DOI. [DOI: 10.1117/1.JBO.19.4.046019]

Keywords: acne vulgaris; thermographic images; co-occurrence matrix; statistical parameters.

Paper 130737RR received Nov. 8, 2013; revised manuscript received Mar. 5, 2014; accepted for publication Mar. 10, 2014; publishedonline Apr. 28, 2014.

1 IntroductionAcne vulgaris is a skin disease affecting more than 80% of theworldwide population with ages between 15 and 44 years old.1,2

Although acne vulgaris is frequently seen as a noncritical der-matological condition, it is highly related with social, psycho-logical, and emotional problems in the patient, such as the onescaused by critical diseases, like asthma, epilepsy, diabetes, andarthritis. The highest effects on the patient are presented whenthe acne is physically located on the face, where it is difficult tohide. One of the effects is the underdevelopment of social skillsin the patient, and consequently interpersonal relationships.3,4

Fortunately, there are several acne treatments available to nota-bly improve the quality of life for patients having this skindisease.

Acne vulgaris as an inflammatory disease, with an excessiveproduction of subdermal fat, modifies the dynamics of the blood-stream, and consequently temperature, on the affected skin zone.Factors responsible for heat variations of the skin might be envi-ronmental, such as temperature, humidity, and speed of wind, ormight be physiological, such as internal body temperature, heatconduction through the tissue, or bloodstream.5 The human bodysurface, as part of a thermal regulation process, requires differentheat interchange levels with the environment. A high percentageof this heat interchange is as electromagnetic radiation within thefar-infrared range, which can be measured through a thermal im-aging camera. Skin damage zones can be located by the heat var-iations sensed through thermal imaging, where the hot areas arecorrelated to skin zones having higher bloodstream. Basically, thethermogram is a skin temperature map, where the increase in the

bloodstream can be identified, indicating skin irritation or lesionin this way.6–9

The use of noninvasive techniques in diagnosing the level ofskin acne is a real challenge; some skin lesions are difficult toidentify using traditional photography because of image satura-tion caused by artificial light, such as flash light. Other tech-niques, having better results, suggest the use of fluorescentphotography for detecting Propionibacterium acne suppressiondue to antibacterial agents, like benzoyl peroxide.10,11 Severalresearch works related to diagnosing acne severity level havebeen reported, most of them are based on quantifying parame-ters, such as types, number, distribution, and density of lesions.Although such diagnosing methods are relatively simple, der-matologists’ experience and subjectivity are stronglyneeded.12–15

The acne severity level (grading) is frequently determinedsubjectively. Different acne grades may be given to the samearea of acne by different dermatologists. However, all reportedtechniques for the evaluation of acne severity level are based onthe number and appearance of acne skin lesions, for whichseverity level is classified into scales. Nowadays, new alterna-tive evaluation techniques for acne severity level are emerging,such as polarized light photography, video microscopy, andmultispectral image analysis, where a direct relationshipbetween skin disease conditions and the spectral characteristicsof reflected light are established.16 Image digital processingmethods are frequently used by such new techniques for taskssuch as segmentation of skin lesion images using binarizationand special filtering algorithms, lesion edges detection, imagecontrast enhancement, and lesion counting and classificationthrough inference algorithms, such as the recently reportedcolor segmentation, K-means clustering, Bayes classifier, maxi-mum entropy classifier, and linear discrimination functions.17–19

*Address all correspondence to: Javier Díaz-Carmona, E-mail: javier.diaz@itcelaya.edu.mx

Journal of Biomedical Optics 046019-1 April 2014 • Vol. 19(4)

Journal of Biomedical Optics 19(4), 046019 (April 2014)

Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Biomedical-Optics on 21 Jul 2020Terms of Use: https://www.spiedigitallibrary.org/terms-of-use

A technique based on thermal images and statistical param-eters for the assessment of acne vulgaris is proposed in thisarticle. The research is focused on finding out a relationship,using first- and second-order statistics, between the acne vulga-ris severity level and the average temperature on the affectedzones. The technique was used to determine the skin thermalbehavior according to its acne severity levels in different stagesof an acne treatment. Infrared images of eight patients diagnosedwith acne vulgaris were captured weekly for 11 weeks, applyingone specific acne treatment until no more improvement in theaffected zone was detected.

2 Materials and MethodsThe study was done on seven patients having acne papulopus-tulosa and one having acne comedones. The patients’ data areshown in Table 1, where gender, age, and acne severity levelbefore, after, as well as in the middle of the study are presented.The acne severity level for each patient was diagnosed by der-matologists using the classification proposed by Hayashi et al.20

Same acne treatment was applied to each patient, consisting aweekly applying of benzoyl peroxide gel, with a concentrationof 1%, on affected skin zone for 5 min; the procedure was com-plemented by applying a water steam flow containing an ozonegas concentration of 0.1%. As it is well known that both sub-stances have antimicrobial and antibacterial properties, beingthe reason why they are frequently used to treat acne.21–23

Each acne treatment session consisted of four stages: (1) skinchemistry exfoliation using alpha-hidroxy acids, (2) saponifica-tion of fat glandules and activity excess using biosulfur and sal-icylic acid, (3) bacterial population control through benzoylperoxide and ozone, and (4) skin inflammation control usingcalendula, chamomile, and aloe vera.

All participants were aware about the study objective andinformed that thermogram capture does not imply any risk tohealth. In order to determine the image capture conditions, sev-eral tests were made previously to the capture section. The RGBacne image of one patient is shown in Fig. 1(a), and the corre-sponding temperature map obtained through the thermogramcapture is presented in Fig. 1(b). From these two images, itis possible to see that the acne zone location agrees with the

thermal skin map, where high-contrasted areas correspond toskin zones affected by acne. Thermal changes on skin zonesdue to the applied acne treatment might be altered by pathologi-cal causes. In order to reduce such causes, patients with someinfection process or some pathology such as diabetes or heartdisease, were not included in the study sample. Besides, preg-nant patients or patients in menstrual period were also excluded.

Before the application of the acne treatment on affected skin,three steps were followed for capturing the thermographicimages. (1) Each patient was relaxed for 30 min in a roomwith a controlled temperature of 25°C (�1°C). (2) In order toobtain an image with improved contrast level, an uniformand thin layer of sunscreen, based on tinanium dioxide crystals,was applied on each patient’s skin24 using a spatula, achievingthermal equilibrium between the skin and the sunscreen canbe achieved. (3) The skin temperature map was obtainedthrough a thermographic camera Flir E45,25 which has a spectralrange from 7.5 to 13 μm and a thermal resolution of 0.2°C, basedon a semiconductor AlGainP diode laser with 1 mW∕635 nm.The camera was placed at a distance of 60 cm from the skin andcalibrated to capture a temperature range between 22°C and35°C with an emissivity coefficient of 0.98.

The above procedure was followed for 11 weeks for eachpatient, always before the acne treatment session, as suggestedby dermatologists to reduce the acne severity level. Some of thepatients showed a significant acne level reduction in just 7weeks. They had a fast response to the applied treatment andfor them only seven thermograms were captured. A treatment

Table 1 Study patients’ data, acne severity levels according toHayashi et al.20

Patientnumber Gender

Age(years)

Acne severitylevel (beforetreatment)

Acne severitylevel (middleof treatment)

Acne severitylevel (aftertreatment)

1 Female 27 Severe Moderate Mild

2 Male 15 Severe Moderate Mild

3 Male 18 Severe Severe Severe

4 Female 23 Severe Moderate Mild

5 Female 22 Moderate Moderate Moderate

6 Male 14 Severe Severe Severe

7 Male 23 Severe Moderate Mild

8 Male 18 Severe Moderate Mild

Fig. 1 Patient’s image samples: (a) RGB and (b) thermograph.

Journal of Biomedical Optics 046019-2 April 2014 • Vol. 19(4)

Padilla-Medina et al.: Assessment technique for acne treatments based on statistical parameters. . .

Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Biomedical-Optics on 21 Jul 2020Terms of Use: https://www.spiedigitallibrary.org/terms-of-use

period of 11 weeks was needed for rest of the patients; hencecapture and analysis of 11 thermographic images were done.The captured thermograms of skin affected by acne were ana-lyzed using first- and second-order statistics. The analyzedimage pixels were chosen as those on an irregular polygonzone covering the affected skin. The same polygon zone wasused for all thermographic images of the same patient.Different polygon zones were applied for different patients,but all having an average zone area of 480 mm2.

2.1 First-Order Statistics

The first-order statistics were computed through the media (m)and standard deviation (σ) of the pixel values on the polygonzone, given respectively as

m ¼ 1

MN

XM−1

i¼0

XN−1

j¼0

fði; jÞ; (1)

and

σ ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi1

MN

XM−1

i¼0

XN−1

j¼0

½fði; jÞ −m�2vuut ; (2)

where i ¼ 0; 1; 2; : : : ;M − 1, and j ¼ 0; 1; 2; : : : ; N − 1 are thecoordinates of each thermogram pixel, M and N represent thesize of analyzed image (M ¼ 320 and N ¼ 240).

2.2 Second-Order Statistics

The second-order statistics were computed through the co-occurrence matrix G, of which element gij represents the num-ber of times that a pixels pair having intensity levels zi and zj arepresent on the image intensities map in a position indicated bythe operator Q, where 1 ≤ i, j ≤ L with L as the number of pos-sible intensities levels. Hence, the concurrence probability ofeach pixels pair is computed as

pi;j ¼gi;jn

; (3)

where n is the total number of pixels pairs satisfying operatorQ.Among the analyzed second-order statistical indices, the

homogeneity parameter (H) was the one that best representedthe acne severity level changes on studied patients. Suchindex is obtained from the probability values pi;j as

26

H ¼XM−1

i¼0

XN−1

j¼0

pi;j

ð1þ ji − jjÞ : (4)

3 ResultsA notable reduction in the acne severity level on five of the eightpatients was achieved by the applied treatment. This reductioncould be confirmed through their average temperature andhomogeneity values computed from the co-occurrence matrix.The average initial and final temperatures, as well as the temper-ature difference for each patient, are presented in Table 2. Adecrease of the average temperature on the acne zone forpatients 1, 2, 4, 6, 7, and 8 was detected. A temperature differ-ence smaller than 2 deg was presented only on patient 6. Higher

final average temperature values were detected on patients 3 and5, which agreed with the increase of acne severity levels on suchpatients. According to t-test, for small samples applied on initialand final temperature values of Table 2, the null hypothesis(Ho: The initial and final temperatures median values areequal) was rejected (α ¼ 0.005, p-value ¼ 0.0028, T-test ¼3.268, Tα ¼ 2.921) and the alternative hypothesis (Ha: Initialtemperature media value is higher than final temperaturemedia value) was accepted.

The thermal behaviors of the average temperature valuesthroughout the treatment for patients 1, 2, 4, 7, and 8 aredepicted in Fig. 2, where each graph point represents the averagetemperature value on the acne zone of each patient.

As it is shown in Fig. 2, only seven thermographic images forpatients 4 and 7 were captured and analyzed, this is because theacne was completely disappeared in the seventh week for bothpatients. Ten and eleven thermographic images were capturedfor the rest of patients. According to Fig. 2, a direct relationshipbetween acne severity level and skin average temperature wasresulted. The lower acne severity level achieved by the treatmentis the greater reduction of the average skin temperature. Theweekly average temperature values of the skin zone affectedby acne for patients with an improvement of acne severitylevel are depicted in Table 3.

As an illustrative example, two thermograms of patient 1 arepseudocolored as shown in Fig. 3. The thermogram at the firstsession, before the acne treatment, is presented in Fig. 3(a). Thethermogram obtained at the tenth week, after the acne treatmentand once a significant reduction of acne severity level was diag-nosed by dermatologists, is shown in Fig. 3(b). From these twoimages, a notable change of the skin temperature map isdetected. A lower temperature value was measured on acneskin zones at the end of the treatment, Fig. 3(b), which agreeswith the achieved acne severity level reduction. The temperaturereduction is also presented on skin area around the acne skinzones. The reason for this global skin temperature decrease isthat a reduction of skin inflammation was resulted by the appliedacne treatment. In this way, a lower face bloodstream as well aproduction drop of fat affected by Propionibacterium acnes bac-teria were achieved. According to Nilsson,6 there is a linear rela-tionship between skin temperature and bloodstream.

Thermal behavior of average skin temperatures for patients 3,5, and 6 (see Table 1) is shown in Fig. 4. The captured skin

Table 2 Average values of initial (T i), final (T f), and difference (ΔT)temperature on acne zone of patients.

Patient T i (°C) T f (°C) ΔT ¼ T i − T f (°C)

1 32.97 28.43 4.49

2 31.47 29.25 2.22

3 31.71 31.74 −0.03

4 32.61 28.88 4.33

5 29.28 30.82 −1.54

6 32.28 30.37 1.91

7 32.44 28.93 3.51

8 32.35 31.12 2.23

Journal of Biomedical Optics 046019-3 April 2014 • Vol. 19(4)

Padilla-Medina et al.: Assessment technique for acne treatments based on statistical parameters. . .

Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Biomedical-Optics on 21 Jul 2020Terms of Use: https://www.spiedigitallibrary.org/terms-of-use

temperatures for patients 3 and 5 at the end of treatment werehigher than the initial ones. According to our hypothesis, thesetemperature results mean an increase of the acne severity level atthe end of the treatment, which was confirmed by the dermatol-ogy results. In the case of patient 6, the acne severity level wasslowly reduced until sixth week, but such achieved level wasincreased in the following four sessions resulting in a highertemperature value at the tenth week.

Thermograms for patient 5 are pseudocolored as shown inFig. 5. The thermogram at the first session, before that acnetreatment, is presented in Fig. 5(a) and the one at eighthweek, after the last treatment session, in Fig. 5(b). As can beseen, there is no significant difference in the average tempera-ture of the acne affected zones in both session images.

The obtained results using first-order statistics were con-firmed with the homogeneity parameter value, computedwith co-occurrence matrix [see Eq. (4)]. The computed homo-geneity values for patients 1, 2, 4, 7, and 8, according to first-order statistics analysis resulted with acne severity improve-ment, are shown in Fig. 6. As can be seen, a reduction

trend of thermogram homogeneity values was obtained. Thefinal homogeneity value is smaller than the initial one forall the cases.

Regarding to the second-order statistical analysis of patientswithout improvement in acne severity level, the behaviors ofthermogram homogeneity values were almost constant alongthe treatment, as can be seen in Fig. 7. For the three cases,the final homogeneity value was greater than the initial one.

4 DiscussionAlthough several diagnosing techniques have been reported fordetermining severity level in inflammatory and noninflamma-tory acnes,10,12,17 visual and touching inspections are themost frequently used techniques by dermatologists when diag-nosing severity level on the affected skin zone. Huamyun andMalik27 proved that it is possible to use multispectral and ther-mal images for classifying acne severity levels; their work isbased on correlation computing of multispectral images.However, such classification still requires a visual analysis ofresulting image and image histogram.

Fig. 2 Thermal behaviors of patients having an improvement with applied treatment.

Table 3 Weekly average temperature values for patients with acne level improvement.

Patient

Week

1 (°C) 2 (°C) 3 (°C) 4 (°C) 5 (°C) 6 (°C) 7 (°C) 8 (°C) 9 (°C) 10 (°C) 11 (°C)

1 32.9 31.0 30.7 31.2 31.7 29.1 30.0 28.7 26.3 28.4 —

2 31.4 30.9 29.7 30.4 29.9 30.0 30.7 31.4 31.1 30.9 29.2

4 32.6 31.1 30.7 33.4 31.8 29.9 28.8 — — — —

7 32.4 31.4 29.9 28.9 30.2 29.9 28.9 — — — —

8 33.3 33.9 33.7 34.0 32.6 29.6 31.0 31.4 30.1 29.7 31.1

Journal of Biomedical Optics 046019-4 April 2014 • Vol. 19(4)

Padilla-Medina et al.: Assessment technique for acne treatments based on statistical parameters. . .

Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Biomedical-Optics on 21 Jul 2020Terms of Use: https://www.spiedigitallibrary.org/terms-of-use

As already reported results show, thermography is a usefultool in determining the acne severity level on skin. This isbecause quantitative information can be obtained from thethermogram, allowing an objective evaluation of the acneseverity level. Although emissivity coefficient of skin may bealtered by the applied sunscreen layer, this emissivity changewas not taken into account in the study analysis because a

small thermal effect is obtained by the thin and applied sun-screen layer, achieving a thermal equilibrium like in the topi-cal-treated skin in Ref. 28. This is based on the small changeof emissivity coefficient value of the skin covered by the sun-screen layer. The computed value of such emissivity coefficientwas 0.92, which is not significantly larger than the emissivity

Fig. 3 Thermograms of patient 1: (a) before and (b) after acnetreatment.

Fig. 4 Average skin temperature values for patients with no acneseverity level improvement.

Fig. 5 Thermograms of patient 5: (a) before and (b) after acnetreatment.

Fig. 6 Thermogram homogeneity values for patients with acneseverity improvement.

Journal of Biomedical Optics 046019-5 April 2014 • Vol. 19(4)

Padilla-Medina et al.: Assessment technique for acne treatments based on statistical parameters. . .

Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Biomedical-Optics on 21 Jul 2020Terms of Use: https://www.spiedigitallibrary.org/terms-of-use

coefficient value of human skin covered without the sunscreenlayer (0.98). The reported results in this article suggest statisticalparameters of media and homogeneity, obtained from the ther-mal maps of acne skin zones, as a helpful tool for determining ifacne severity level is reduced or not throughout the use of onespecific acne treatment. Such statistical parameters might bemainly helpful for efficiency assessment in early stages ofacne treatments, when neither visual nor touching results areavailable, approximately at fourth or sixth week.

Among all second-order statistical indices computed fromcaptured thermographic skin images, the homogeneity param-eter was the one with the best results in quantitatively estimatingthe acne severity level along the applied acne treatment. Thehigh-homogeneity values were obtained for patients withhigh-acne severity levels, and the smallest values for patientswith low and moderate acne severity levels. From obtainedresults, an explanation to this homogeneity value behavior isthat image pixels values are locally homogeneous on skinzone affected by acne, as long as acne severity level is reducedthe pixels values are scattered, reducing in this way their homo-geneity as a consequence of the infrared radiation dispersionreflected on the human skin.

The use of the first-order statistics, based on the mediaparameter computing from captured images, allowed the globalevaluation of temperature variations on skin affected by acne.On the other hand, local temperature variations were possibleto be detected by using the second-order statistics and the homo-geneity parameter value. Therefore, it is possible to asseveratethat the average temperature value on affected skin zone isdecreased, the temperature gradient value of local skin zonesis incremented, and the skin behaves as a healthy one withan infrared emission similar to a normal distribution.29

5 ConclusionsStatistical analysis results of thermal images captured frompatients’ skin having acne vulgaris along a specific treatmentare presented in this article. The project goal was focused tofind out a possible relationship between the acne severitylevel and the obtained statistical data along the treatment stages.

The computing of several statistical parameters was done forthermal images captured from eight patients in a period of time

of 7 and 11 weeks. According to the obtained results, the first-and second-order statistical parameters media and homogeneity,respectively, showed a direct relationship to the acne severitylevel. In this way, such statistical parameters may be used asa quantitative assessment tool for acne treatments efficiencymainly in early stages, when neither visual nor touching resultsare available, approximately at the fourth or sixth week. Theproposed method might be extended as future work to applica-tions on treatment assessment of other skin diseases like psoria-sis, vitiligo, or even skin burns.

AcknowledgmentsThe authors would like to thank the Angel Care Dermal-Cosmetic Institute for the facilities in the project development,and CONACyT and DGEST for the financial support.

References1. W. Hongcharu et al., “Topical ALA-photodynamic therapy for the treat-

ment of acne vulgaris,” J. Invest. Dermatol. 115(2), 183–192 (2000).2. P. M. Friedman et al., “Treatment of inflammatory facial acne vulgaris

with the 1450-nm diode laser: a pilot study,” Dermatol. Surg. 30(2),147–151 (2004).

3. R. Alharithy, “Adolescent’s acne: scarring inside out!,” J. Saudi Soc.Dermatol. Dermatol. Surg. 15(2), 43–46 (2011).

4. L. E. Barness et al., “Quality of life measures for acne patients,”Dermatol. Clin. 30(2), 293–300 (2012).

5. C. M. Burgess, Cosmetic Dermatology, Springer, Heidelberg, Germany(2005).

6. A. L. Nilsson, “Blood flow, temperature, and heat loss of skin exposedto local radioactive and convective cooling,” J. Invest. Dermatol.88(5), 586–593 (1987).

7. N. A. Diakides and J. D. Bronzino, Medical Infrared Imaging, CRCPress, New York (2008).

8. Y. Houdas and E. F. J. Ring, Human Body Temperature ItsMeasurement and Regulation, Plenum Press, New York (1982).

9. S. J. Yoon, S. C. Noh, and H. H. Choi, “Thermographic diagnosis sys-tem and imaging algorithm by distributed thermal data using simpleinfrared sensor,” Curr. Appl. Phys. 10(2), 487–497 (2010).

10. L. C. Lucchina et al., “Fluorescence photography in the evaluation ofacne,” J. Am. Acad. Dermatol. 35(1), 58–63 (1996).

11. A. Pagnoni et al., “Digital fluorescence photography can assess the sup-pressive effect of benzoyl peroxide on propionibacterium acnes,” J. Am.Acad. Dermatol. 41(5), 710–716 (1999).

12. J. A. Witkowski and H. M. Simons, “Objective evaluation of demethychlortetracycline hydrocholoride in the treatment of acne,” J. Am. Med.Assoc. 196(5), 397–400 (1966).

13. C. H. Cook, R. L. Centner, and S. E. Michaels, “An acne grading methodusing photographic standards,” Arch. Dermatol. 115(5), 571–575 (1979).

14. D. A. Winstanley and N. S. Uebelhoer, “Future considerations in cuta-neous photomedicine,” Sem. Cutaneous Med. Surg. 27(4), 301–308(2008).

15. J. K. Tan, K. Fung, and L. Bulger, “Reliability of dermatologists in acnelesion counts and global assessments,” J. Cutaneous Med. Surg. 10(4),160–165 (2006).

16. M. Yamaguchi et al., “Multispectral color imaging for dermatology:application in inflammatory and immunologic diseases,” in Societyfor Imaging Science and Technology (IS&T), Proc. 13th ColorImaging Conf., Arizona, pp. 52–57 (2005).

17. R. Ramli et al., “Acne analysis, grading and computational assessmentmethods: an overview,” Skin Res. Technol. 18(1), 1–14 (2012).

18. H. Fujii et al., “Extraction of acne lesion in acne patients frommultispectralimages,” inProc.Ann. Int.Conf. IEEEEng.Med.Biol. Soc., Engineering inMedicine and Biology Society (EMBS), pp. 4078–4081 (2008).

19. A. S. Malik et al., “Novel techniques for enhancement and segmentationof acne vulgaris lesions,” Skin Res. Technol. (2013).

20. N. Hayashi, H. Akamatsu, and M. Kawa-Shima, “Establishment ofgrading criteria for acne severity,” J. Dermatol. 35(5), 255–260(2008).

Fig. 7 Thermogram homogeneity values for patients with no acneseverity improvement.

Journal of Biomedical Optics 046019-6 April 2014 • Vol. 19(4)

Padilla-Medina et al.: Assessment technique for acne treatments based on statistical parameters. . .

Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Biomedical-Optics on 21 Jul 2020Terms of Use: https://www.spiedigitallibrary.org/terms-of-use

21. M. A. Herane, “Actualización terapéutica en acne vulgaris,” Rev. Ofic.Soc. Latin. de Dermat. Pediat. 3(1), 5–19 (2005).

22. M. Sharma and J. B. Hudson, “Ozone gas is an effective and practicalantibacterial agent,” Am. J. Infec. Control 36(8), 559–563 (2008).

23. L. Re et al., “Ozone therapy: clinical and basic evidence of its thera-peutic potential,” Arch. Med. Res. 39(1), 17–26 (2008).

24. C. Villaseñor-Mora, F. J. Sanchez-Marín, and M. E. Garay-Sevilla,“Contrast enhancement of mid and far infrared images of subcutaneousveins,” Infrared Phys. Tech. 51(3), 221–228 (2008).

25. FLIR Systems, “Thermal CAM E45,” February 6 2006, http://support.flir.com/DocDownload/Assets/36/English/1558015$a155.pdf (7 April2014).

26. R. C. González and R. E. Woods, Digital Image Processing, 3rd ed.,Prentice Hall, New Jersey (2008).

27. J. Huamyun and A. S. Malik, “Multispectral and thermal images for acnevulgaris classification,” in Proc. Natl. Post. Conf. (NPC), pp. 1–4,Universiti Teknologi Petronas (UTP) (2011).

28. V. Bernard et al., “Infrared camera assessment of skin surface temper-ature–effect of emissivity,” Phys. Med. 29(6), 583–591 (2013).

29. N. Ludwing et al., “Skin temperature evaluation by infrared thermog-raphy: comparison of image analysis methods,” Infrared Phys. Tech. 62,1–6 (2014).

J. Alfredo Padilla-Medina received his PhD degree in optics fromOptics Research Center, Leon, Gto., México (2003), MS degree inelectrical engineering from Univeristy of Guanajuato, Salamanca,Gto., México (1995), and BS degree in electronics engineeringfrom Technological Institute of Celaya (ITC), Celaya, Gto, México(1993). Currently, he is a full-time professor/researcher at theElectrical and Electronics Engineering Department at ITC. Hisresearch interests include digital visible/infrared image processing,the human visual system, and fuzzy logic.

Francisco León-Ordoñez received his MS degree in electronicsengineering from Technological Institute of Celaya, Celaya, Gto,México (2009), and BS degree in electronics engineering fromTechnological Institute of San Juan del Rio, Qro., México in(2006). Currently, he is a professor at the Chemistry EngineeringDepartment of Tecnological University of San Juan del Rio. Hisresearch activities are focused on teaching and industrial counseling.

Juan Prado-Olivarez received his PhD degree in instrumentation-microelectronics from Université Henri Poincaré, Nancy I, France(2006), MS degree in electrical engineering from University ofGuanajuato, Salamanca, Gto., México (2000), and BS degree in elec-tronics engineering from Technological Institute of Celaya (ITC),Celaya, Gto., México (1993). Currently, he is a full-time professor/researcher in the Electrical and Electronics EngineeringDepartment in ITC. His research interests include bioimpedancespectroscopy for tissues characterization and dielectric study forassessment of human patologies.

Noe Vela-Aguirre received his PhD degree in bioengineering withemphasis in biolectronics from Polytechnique University ofValencia, Spain (2008), MS and BS degrees in electrical engineeringfrom University of Guanajuato, Salamanca, Gto, México (1987), and(1985), respectively. Currently, he is a full-time professor/researcherat the Electrical and Electronics Engineering Department ofTechnological Institute of Celaya, Celaya Gto., México. His researchinterests include monitoring and processing of biological signals.

Agustin Ramirez-Agundis received his PhD in digital systemsdesign from Polytechnique University of Valencia, Spain (2008),MS and BS degrees in electrical engineering from University ofGuanajuato, Salamanca, Gto, México (1977), and (1975), respec-tively. Currently, he is a full-time professor/researcher in theElectrical and Electronics Engineering Department of TechnologicalInstitute of Celaya, Celaya Guanajuato, México. His research inter-ests include automatization of industrial test machines and hardwaresystems development based on reconfigurable devices.

Javier Diaz-Carmona received his PhD and MS degrees in electron-ics from National Institute of Astrophysics Optics and Electronics,Puebla, México (1997) and (2003), respectively, and a BS degreein electronics engineering from Technological Institute of Celaya(ITC), Celaya, Gto., México (1990). Currently he is a full-timeprofessor/researcher in the Electrical and Electronics EngineeringDepartment in the ITC. His research interests are focused ondigital signal processing and reconfigurable hardware embeddedsystems.

Journal of Biomedical Optics 046019-7 April 2014 • Vol. 19(4)

Padilla-Medina et al.: Assessment technique for acne treatments based on statistical parameters. . .

Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Biomedical-Optics on 21 Jul 2020Terms of Use: https://www.spiedigitallibrary.org/terms-of-use